Consecutive runs of a Functional Test do not populate screenshot differences in the Screen Comparison Tab

Consecutive runs of a Functional Test do not populate screenshot differences in the Screen Comparison Tab

Discovered during the verification of [Link Removed], that issue was closed due to the setup not quite being complete

also occurs in //UE4/Release-4.22 at CL 7053642

Setup:

• Have Git installed

Execution:

1. Create and open a new Blueprint First Person template project
2. Using the modes panel, add a Screenshot Functional Test actor to the level
   • Move the actor so an object can be easily framed/moved/placed in front of it
3. Save the Level
4. Open Window > Developer Tools > Session Frontend
5. Select the "Automation" tab & Search for "ScreenshotFunctionalTest"
6. Select the check box next to the ScreenshotFunctionalTest & Select "Start Tests"
7. After the test is complete, focus the "Screen Comparison" tab
8. Select "Source Control" > Connect to Source Control from the Level Editor toolbar
9. Select "Git" from the provider dropdown
10. "Initialize project with Git" & "Accept Settings"
11. Refocus the Screen Comparison tab in Session Frontend and "Add New!"
12. Without closing the Session Frontend, add a cube to the ScreenshotFunctionalTest's view
13. Save the level
14. In the Session Frontend, Select the Automation Tab and re-run the Test
15. After the test fails due to updated screenshot, Focus the Screen Comparison tab

Result:
No Ground Truth/Delta/Incoming pictures, even though the screenshots exist on disk and error is produced in the Output Log and Messages

Expected:
Ground Truth/Delta/Incoming entry for a failing test

Workaround:
Close and re-open SessionFrontend, the screenshots will be populated